ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: 1 2 [All]   Go Down

Author Topic: JBL VTX  (Read 8399 times)

Rasmus Rosenberg

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 55
JBL VTX
« on: May 28, 2012, 08:45:17 AM »

Hello,
Has anyone measured the the JBL VTX? I tried at a demo and a friend had it on a festival this weekend outdoor gigs almost no wind. 2 different setups (9 and 12 boxes) and not my own measurement setup, but same thing happened. With Smaart it was very hard to get the internal (Ref) delay right, ( I never seen any thing like it ). Verified the measurement rig both times and couldn't find any faults. What happened is when you measure more than 4 boxes, smaart will not lock on (find the delay time). In the impulse response the more boxes you add the more little spikes before and after the "biggest" spike, but its almost a random number, not scaleable with box count (only the more boxes the more spikes). About maybe 12-15 spikes if not more. I have had speakers/linearrays measure "funky" before (where you see small spikes rise the more boxes you add, maybe 2-5 very little spikes near the biggest) , but no way near this level and time spread wise. Tried with a "qualified guessed" delay time and could in one instance get a useful measurement up to about 6khz, but on the others not higher than about 1-2khz. In both instances the rigs sounded good in the high end, no funky phasing sound or any hearable clues to something was wrong. Sorry for the vauge explanation only had limited measurement time at the demo, and that I don't have any traces, as it would be much easier to comment on. 
Any way got me puzzled, because the same trend happened on two different setups,  any tips or clues would be nice.
Rasmus

 
Logged

Paul Tucci

  • SR Forums
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 371
Re: JBL VTX
« Reply #1 on: May 28, 2012, 11:45:44 AM »

Hello,
Has anyone measured the the JBL VTX? I tried at a demo and a friend had it on a festival this weekend outdoor gigs almost no wind. 2 different setups (9 and 12 boxes) and not my own measurement setup, but same thing happened. With Smaart it was very hard to get the internal (Ref) delay right, ( I never seen any thing like it ). Verified the measurement rig both times and couldn't find any faults. What happened is when you measure more than 4 boxes, smaart will not lock on (find the delay time). In the impulse response the more boxes you add the more little spikes before and after the "biggest" spike, but its almost a random number, not scaleable with box count (only the more boxes the more spikes). About maybe 12-15 spikes if not more. I have had speakers/linearrays measure "funky" before (where you see small spikes rise the more boxes you add, maybe 2-5 very little spikes near the biggest) , but no way near this level and time spread wise. Tried with a "qualified guessed" delay time and could in one instance get a useful measurement up to about 6khz, but on the others not higher than about 1-2khz. In both instances the rigs sounded good in the high end, no funky phasing sound or any hearable clues to something was wrong. Sorry for the vauge explanation only had limited measurement time at the demo, and that I don't have any traces, as it would be much easier to comment on. 
Any way got me puzzled, because the same trend happened on two different setups,  any tips or clues would be nice.
Rasmus

Rasmus,

I think the conclusion you have reached and are trying to support is false. That is, a valid measurement is unobtainable with more than a few boxes. That thinking directly contradicts my career.
Odds are that operator error is the culprit. An unfamiliar rig without enough time to investigate can create some uncertainties.
In the interest of learning I'll offer up my clues. Verifying the measurement rig is a good place to start. Is there any bleed from Ref to Msrmt channels? That will contaminate a measurement quickly. Is the FFT size in the IR measurement large enough? Bigger yields better data. Are you averaging enough to help the S/N ratio? More averaging yields a higher S/N ratio. Asking yourself " Do I even have hope of getting a good measurement?"  ie Measuring during the show with a significant level bleeding off the stage into the measurement mic will get in the way, as would the other side of the PA being on also. You say you were outdoors so reflections were of little concern, that's one variable accounted for.

Tip Time:
. Realize you DON'T need an impulse response to get a good measurement. (SAY WHAT?)
The reason we offset the reference signal to the measurement signal is to make our comparison of the reference signal AT the same time  The clue that we have the correct offset time is a flat phase trace in the TF. Typically this occurs in the high end, but it need not.
One could purposefully use an offset time that mucked up the high end but results in a flat trace elsewhere. 
That opens up a whole world of possibilities like guestimating the offset time in your situation and then being able to verify that you're correct by seeing a stable, flat phase trace in that area of interest.

We live under the illusion that all the frequencies arrive at our ears (or the measurement mic) at the same time. If that were the case flat phase traces would be prevalent and might even show up on manufacturers' literature. Can I get an amen Langston? Knowing the implications of manipulating the offset time changes everything.

A bit rambly, but hopefully helpful

PT
Logged

Langston Holland

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 82
Re: JBL VTX
« Reply #2 on: May 28, 2012, 02:01:21 PM »

That kind of rambling requires years of measurement experience. Amen. :)
Logged
God bless you and your precious family - Langston

Ivan Beaver

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 9534
  • Atlanta GA
Re: JBL VTX
« Reply #3 on: May 28, 2012, 04:32:09 PM »


We live under the illusion that all the frequencies arrive at our ears (or the measurement mic) at the same time. If that were the case flat phase traces would be prevalent and might even show up on manufacturers' literature. Can I get an amen Langston? Knowing the implications of manipulating the offset time changes everything.

A bit rambly, but hopefully helpful

PT
But when they DO arrive at the same time-and at the same time throughout the coverage pattern-it DOES make a difference.

Especially outdoors and with wind involved.
Logged
A complex question is easily answered by a simple-easy to understand WRONG answer!

Ivan Beaver
Danley Sound Labs

PHYSICS- NOT FADS!

Paul Tucci

  • SR Forums
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 371
Re: JBL VTX
« Reply #4 on: May 28, 2012, 04:58:53 PM »

But when they DO arrive at the same time-and at the same time throughout the coverage pattern-it DOES make a difference.

Especially outdoors and with wind involved.

Ivan,

That's one of the characteristics that we can strive for as system engineers. Linear phase throughout the coverage area.

If you can figure out how to make sound arrive at the front row AND the back row of the coverage area simultaneously, please enlighten me! :)

PT
Logged

Langston Holland

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 82
Re: JBL VTX
« Reply #5 on: May 28, 2012, 05:23:44 PM »

Quote from: Paul
If you can figure out how to make sound arrive at the front row AND the back row of the coverage area simultaneously, please enlighten me! :)

At your service! Band goes in the ceiling!
Logged
God bless you and your precious family - Langston

Paul Tucci

  • SR Forums
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 371
Re: JBL VTX
« Reply #6 on: May 28, 2012, 05:42:19 PM »

At your service! Band goes in the ceiling!

I just have the sense you've been waiting for the opportunity to use that photograph.
PT
Logged

Rasmus Rosenberg

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 55
Re: JBL VTX
« Reply #7 on: May 29, 2012, 07:32:22 AM »

Hey Paul,
Thanks for commenting very helpful. I think I'm homing in on the issue, hopefully i can get some bench testing time soon. While it does not support why it happened on two different measurement setups, what i think was wrong on the setup i saw was:
Cross talk at the Ref channel. At least looked a lot like this emulated picture.
https://www.dropbox.com/sh/jwgay8jfhxjbize/z0DXlgY0-Y#f:Crosstalk.jpg

mvh
Rasmus

Edit: The link is for dropbox there is a pic of background noise, not related to this thread.
« Last Edit: May 29, 2012, 07:40:48 AM by Rasmus Rosenberg »
Logged

Paul Tucci

  • SR Forums
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 371
Re: JBL VTX
« Reply #8 on: May 29, 2012, 01:00:15 PM »

Look at all that comb filtering in the TF!  I've seen that when the reference signal takes two different paths to get to a bus output. ie through a subgroup AND a matrix.
Equal level through each path but with different throughput times = comb filters when summed.

PT
Logged

Rasmus Rosenberg

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 55
Re: JBL VTX
« Reply #9 on: May 29, 2012, 01:30:53 PM »

Look at all that comb filtering in the TF!  I've seen that when the reference signal takes two different paths to get to a bus output. ie through a subgroup AND a matrix.
Equal level through each path but with different throughput times = comb filters when summed.

PT

Yes, that would also not show up in a loopback test of the measurement rig alone. Also while you can see it, on the capture spectrum measurement of the ref channel I just uploaded, it was hard to see that combing when the trace ran live. Top is contaminated bottom is the internal generator directly.

 
« Last Edit: May 29, 2012, 01:41:00 PM by Rasmus Rosenberg »
Logged

Ivan Beaver

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 9534
  • Atlanta GA
Re: JBL VTX
« Reply #10 on: May 29, 2012, 07:53:36 PM »

At your service! Band goes in the ceiling!
Not to rain on your parade-but as long you want ot "ignore" some basic fundamental ideas- your idea would work.

What about the "unseen" chair in the middle?  The sound would get to it first????????

Yeah-Yeah-just like a lot of other things in audio that people just "choose" to ignore--------------------------

Just kidding.  And yet again-it is not that simple.
Logged
A complex question is easily answered by a simple-easy to understand WRONG answer!

Ivan Beaver
Danley Sound Labs

PHYSICS- NOT FADS!

ProSoundWeb Community

Re: JBL VTX
« Reply #10 on: May 29, 2012, 07:53:36 PM »


Pages: 1 2 [All]   Go Up
 



Site Hosted By Ashdown Technologies, Inc.

Page created in 0.036 seconds with 24 queries.